The popularity of extreme sports (also referred to as “action sports” or “adventure sports”) is steadily growing. In particular, sports such as snowboarding, skateboarding, free skiing, surfboarding, skydiving, wingsuit flying, bicycle motocross (BMX), and others are becoming (or are currently) mainstream sports. Such sports are increasingly being covered by various media organizations and some competitions (such as the X-Games) are devoted solely to extreme sports.
In many traditional sports, competitions are measured and judged using objective measures such as scores (e.g., the number of runs in a baseball game or points in a basketball game) or times (e.g., the time for a runner or downhill skier to cross a finish line). In many extreme sports, by contrast, athletes compete by performing various athletic maneuvers (or “tricks”) such as jumps, flips, rotations, and the like. As an example, a description of common tricks that are performed in snowboarding can be found at http://en.wikipedia.org/wiki/List_of_snowboard_tricks and a description of common skateboarding tricks at http://skateboardingtrickslist.com.
Many athletic maneuvers in extreme sports are complex and/or performed very quickly. Accordingly, athletic maneuvers performed in action sports competitions are often evaluated (by judges and spectators) based on a variety of subjective factors, including an individual's personal perception of the difficulty or aesthetics of a particular athletic maneuver. This often leads to problems in identifying different athletic maneuvers, as well as judging or rating such maneuvers in competition. In ski jump competitions, for example, the movement of the jumper may be so fast that spectators cannot determine the number of rotations or flips the jumper performs.
Some conventional systems attempt to measure fast rotations with inertial sensors. Such measurements are typically made relative to the sensor axis but can be related to the absolute axis as well. However, such conventional systems do not make sensor measurements understandable to the sport participants and spectators. Conventional systems are unable to automatically recognize and measure action sport “tricks,” and are thus unsuitable for virtual competitions, leaderboards, and social networks dedicated to the action sport participants.
Additionally, fans of traditional and extreme sporting events often record videos of such events using video cameras, their smart phones, or other video-capturing devices. This provides a large number of possible video sources for various events, which can be further supplemented by video drones and other systems.
While such recording is often not professional, it provides a vast amount of video data that could be mined to create a unique viewing experience. However, the very nature of having multiple video sources, many of which may be unattended and/or uncoordinated, creates a serious challenge for data selection and editing. Embodiments of the present disclosure address these and other issues.